Four decision simulations for policy education, professional learning and workshops. Each case puts participants in a real decision with incomplete evidence, competing stakeholders and a deadline, and asks them to reach a defensible answer.
Whose Cloud, Whose Rules? — public-sector AI procurement and sovereignty: whether, and on what terms, to authorise a government generative-AI platform hosted by a foreign supplier
The Risk Score Before the Court — justice and due process: whether courts may use a machine-learning risk score, and under what safeguards and rights of challenge
Credit Without Trust — small-business finance and inclusion: whether an AI-enabled credit-scoring pilot may influence lending decisions
The Advisory Last Mile — agricultural advisory and local knowledge: whether to continue, redesign, restrict or stop a service after early results
This edition unifies what were previously two separate volumes. Part I holds the four participant packets — the decision and mandate, case narrative and exhibits, stakeholder submissions, role cards written without model arguments, and decision worksheets — and may be shared with participants in full. Part II holds the four facilitator guides — delivery logic and timing, inject instructions, defensible decision families, common misconceptions and the interventions that answer them, moderation anchors and release audits. Part I occupies one continuous page range so it can be printed or sent on its own.
All four cases test the same seven judgements — public problem before technology; authority and institutional responsibility; evidence, uncertainty and proportionality; rights, inclusion and meaningful remedy; implementation capacity and non-AI alternatives; vendor power, data governance and exit; monitoring, material change, pause and stop conditions — so a participant who has run one case recognises the structure of the next. Each runs as a standard workshop of 150 minutes or a rapid lab of 60 minutes.
The cases were developed from the eight capstone projects submitted by Cohort 1 of the AI Literacy Fellowship for African Policymakers, written by forty-one fellows and published in full in the companion volume. The Republic of Ubuntia, its institutions, people, laws, data and events are fictional unless an external source is expressly cited.
Published by the OpenSchool Initiative, Abuja, Nigeria. First open publication edition, 2026. Licensed CC BY 4.0.
This is an open educational and policy-learning resource. Anyone adapting a case for a real jurisdiction must replace the fictional authorities, laws, figures and constraints with verified local material, engage affected groups and relevant specialists, and disclose material changes.